Two location markers connected across a distance

I'm a local developer, so treat this with the scepticism it deserves — and then check it against your own numbers, because offshore development genuinely wins in more cases than people in my position usually admit.

The honest version is that the hourly rate tells you almost nothing. What decides the outcome is how much specification you can write, how much management you can supply, and how tolerant the project is of a slow feedback loop.

The rate comparison, and why it misleads

Indicative 2026 rates for small-business application work.
OptionHourlyWhat the rate includes
Offshore marketplace freelancer$15–$40Code against a specification you supply. Little else.
Offshore agency$30–$70 Developers plus a project manager and some QA. Specification still largely yours.
Local independent developer$100–$175 Requirements, build, testing and deployment from one person who talks to you directly.
Domestic agency$150–$300 A team, plus account management, plus the overhead of both.

A 4× rate difference looks decisive until you account for the hours that don't appear on the invoice. The offshore rate assumes someone has already decided what to build. If that person is you, your time is part of the cost.

The same project — a single-workflow replacement, roughly 150 developer-hours of actual coding.
Cost elementOffshore agencyLocal independent
Development hours billed190 × $50 = $9,500110 × $120 = $13,200
Your time writing specifications40 hrs10 hrs
Your time in review and rework loops30 hrs10 hrs
Your time valued at $75/hr$5,250$1,500
True total$14,750$14,700
The invoice is not the costSame project: a single-workflow replacement, ~150 hours of coding
Invoiced to you Your own time, valued at $75/hr
$0 $3.8k $7.5k $11.3k $15k Offshore agency · Invoiced to you: $9,500 $9,500 Offshore agency · Your own time, valued at $75/hr: $5,250 $5,250 Local independent · Invoiced to you: $13,200 $13,200 Local independent · Your own time, valued at $75/hr: $1,500 $1,500 Offshore agency Local independent

The rate difference is real. So is the specification and review time it transfers to you. Which one is cheaper depends on whether your attention or your cash is the scarcer resource.

These come out level — which is the actual finding. Neither option is obviously cheaper for a typical small-business project. What differs is what kind of cost you pay: money, or your own attention.

Where offshore genuinely wins

  • The specification already exists. If you have designs, a written spec, or an existing system being ported, the biggest advantage of proximity disappears and the rate difference becomes real.
  • The project is large enough to need a team. Above roughly 500 hours, an agency's structure starts earning its overhead and one local developer becomes a bottleneck.
  • You have a technical person in-house. Someone who can review work and answer questions precisely removes most of the communication penalty.
  • The work is well-bounded and repetitive. Building forty similar screens is exactly the kind of work that parallelises well and needs little judgement.

Where it usually doesn't

  • You can't fully specify what you want. Most small businesses can't, and this is not a failing — the requirement genuinely emerges through conversation with someone who understands the business. A relationship where every ambiguity costs a day of round-trip is expensive in a way no rate card shows.
  • The project is small. Under 100 hours, coordination overhead swamps the rate advantage.
  • The work needs to be seen in place. Anything involving your physical premises, your hardware, your network or your staff being trained in a room.
  • You need one accountable person. Agency staffing changes; the developer who understood your business in month two may not be there in month six.

The risks people actually hit

These apply to both, and they matter more than geography.

  • Code you don't own. Check the contract, not the sales call. If the repository is in their account, you have a subscription with extra steps.
  • Nobody left who understands it. Turnover ends more projects than incompetence does. Ask what happens when your developer leaves, and get the answer in writing.
  • Time zones and the feedback loop. A 10-hour gap turns a five-minute clarification into a one-day delay. Over a project that's weeks, and it's the cost people consistently underestimate.
  • Fixed price with a vague scope. The cheapest quote against the loosest specification is the most reliable predictor of a project that ends badly, at any rate and in any country.

Questions to ask whoever you're considering

  1. Who owns the source code, and whose account is the repository in?
  2. Can I see something running in week two, or do I wait until the end?
  3. Who exactly will write this, and will they still be here in month six?
  4. What's excluded from this quote? (Listen for migration and integrations.)
  5. What happens after launch, and what does that cost?
  6. Can you show me something you built and still run today?

That last one is worth more than the rest combined. Anyone can show you a portfolio; far fewer can show you something they've operated for years.

The summary I'd give a friend

If you know exactly what you want and have someone to check the work, offshore is a reasonable way to buy more hours for the money. If the requirement is still fuzzy, the project is small, or you want one person who understands your business and answers the phone, local costs about the same once your own time is counted — and it costs a great deal less in aggravation.

Before either, though, make sure the project is worth doing at all. The build-vs-buy checklist and the break-even calculator will tell you that in about ten minutes, and they're a cheaper place to find out than a quote process.

Want this done for you? Send me your last month of software invoices and I'll tell you what's worth replacing, what isn't, and roughly what a replacement would cost. It's free, and "keep paying for it" is a normal answer.

Get a free subscription audit